The study protocol was preregistered in PROSPERO (CRD42024564331). The manuscript was structured following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist [20], and the checklist is presented in Multimedia Appendix 1.

The initial comprehensive literature search for this meta-analysis was conducted in February 2024 and updated in August 2024 to capture any newly published studies. The search spanned 4 major English-language databases: PubMed, Embase, Cochrane Library, and Web of Science. Keywords derived from relevant articles were used, including terms such as telemetry, digital health, e-consultation, telemonitoring, smartphone technology, online communication, and digital health technology. Details of the search strategy are provided in Multimedia Appendix 2.

Eligible studies were required to be RCTs published in English, focusing on pregnant women aged 18 years and older. Digital health interventions, such as phone calls, text messages, and interactive apps (eg, YouTube, Twitter, and WeChat), were implemented in the intervention group, while the control group received standard care. The studies evaluated either maternal outcomes, such as GWG and pregnancy complications, or neonatal outcomes, such as preterm birth and SGA, defined as a birth weight at or below the 10th percentile for gestational age.

Studies were excluded if they did not include a control group, involved both intervention and control groups receiving digitally mediated treatments, failed to report the desired outcomes, measured them only postpartum, or lacked accessible full-text articles or usable data.

Following the removal of duplicates, 2 researchers (JW and NT) independently screened the remaining articles by evaluating their titles and abstracts, excluding those that were irrelevant. The full texts of studies identified as potentially relevant were retrieved and further assessed to determine their eligibility for inclusion. In cases where discrepancies arose between the two researchers, these were resolved through discussion or, if necessary, by consulting a third researcher to achieve consensus.

For articles selected for further analysis, JW and NT used a standardized data extraction worksheet developed in Microsoft Excel 2016. The extracted data encompassed key study characteristics, including the year of publication, authors, country, sample size, inclusion and exclusion criteria, maternal and neonatal health outcomes, gestational age at enrollment, prepregnancy maternal BMI, type of digital applications, specific interventions, control measures, duration of intervention, high-risk factors of participants, and the effects of digital care on maternal and neonatal health. Any disagreements during the data extraction process were resolved through iterative discussions between the authors until a consensus was reached.

The bias of the RCTs included in this meta-analysis was evaluated using the Revised Cochrane risk-of-bias tool for randomized trials (RoB 2) [21]. JW and NT independently assessed the risk of bias for each included study. Discrepancies between the two reviewers were resolved through iterative discussions, revisiting study details, and assessment criteria to reach a consensus. If disagreements persisted, a third reviewer provided an independent evaluation, with the final decision based on majority agreement.

Data analysis for the meta-analysis was performed using Stata (version 15.1; StataCorp). Effect sizes were calculated and presented as forest plots to facilitate quantitative synthesis. Standardized mean differences (SMDs) were used for continuous variables, while odds ratios (ORs) were applied for dichotomous outcomes. The choice between fixed-effect and random-effects models was determined by the level of heterogeneity, with I² values above 50% indicating substantial heterogeneity; a random-effects model was used for I²>50%, while I²≤50% warranted a fixed-effect model. A P value of <.05 was considered statistically significant. To assess the robustness of the synthesized results, sensitivity analyses were conducted by sequentially excluding each study and reanalyzing the data, as well as by restricting the analysis to studies with a low risk of bias. To assess a potential publication bias, funnel plots were used, and the Egger regression test was used to calculate the publication bias (Multimedia Appendix 3).

A total of 19,936 studies were retrieved from the 4 databases, and 1 additional article was identified through a manual search. After removing duplicates, 11,606 studies remained for further evaluation. Titles and abstracts were screened, and 69 studies were deemed relevant. Following a detailed review of the full texts, 42 RCTs met the inclusion criteria and were incorporated into the meta-analysis. Figure 1 shows the complete screening process.

This meta-analysis includes data from 42 RCTs involving a total of 148,866 participants. Its primary focus is to evaluate the impact of digital health interventions on maternal and neonatal outcomes, particularly among populations with gestational diabetes mellitus (GDM) and other high-risk pregnancy factors. The studies, conducted between 2007 and 2024, were geographically distributed across Asia (13/42 studies, 31%), Europe (13/42 studies, 31%), North America (9/42 studies, 21%), Australasia (6/42 studies, 14%), and Africa (1/42 study, 3%). The majority of studies were published in the United States (9/42 studies, 21%), Australia (6/42 studies, 14%), and China (6/42 studies, 14%).

Digital health interventions used three primary delivery modalities: (1) mobile devices, including smartphones and tablets; (2) website-based platforms; and (3) mobile apps incorporating social software (eg, Facebook [Meta Platforms], Zoom [Zoom Communications, and WeChat [Tencent Holdings Limited]) and other digital health tools. Table 1 presents the detailed characteristics of the 42 RCTs [22-63].

A total of 25 RCTs included in this review, focusing on GWG, were evaluated using the RoB 2 tool. Among these, 11 studies were determined to have a low risk of bias, 3 were identified as high risk, and 11 presented some concerns regarding potential bias. Studies that focused on secondary outcomes, such as neonatal health, were not included in this assessment, as the RoB 2 evaluation was specifically applied to studies addressing the primary outcome of GWG. Figure 2 [22-25,27-29,31,32,34,42-46,48,50,52,54-56,58,61-63] shows the risk-of-bias assessment.

This meta-analysis revealed that digital health interventions significantly improved excessive GWG, reduced miscarriage and preterm birth incidence, and enhanced neonatal outcomes. However, the benefits for women with insufficient weight gain were limited, and an increased rate of infants who are SGA was observed.

This review evaluated the impact of digital health interventions on maternal and neonatal outcomes, with a particular focus on GWG. The findings revealed that digital interventions significantly improved GWG management among pregnant women. Compared to the control group, the intervention group showed a notable decrease in women exceeding recommended weight gain and increased adherence to the IOM GWG guidelines (P<.05). These results align with previous meta-analyses by Islam et al [64] and He et al [16]. However, the reduction in weight gain observed in our study (–0.145 kg) was less pronounced than the reductions reported by Islam et al [64] (–1.07 kg) and Antoun et al [65] (–1.99 kg). This discrepancy may be attributed to differences in the types and intensity of digital interventions and participant characteristics.

Subgroup analysis further revealed that significant GWG reductions were predominantly observed in participants who were overweight or obese rather than those with GDM. This may be explained by the intensive medical management typically provided to patients with GDM, which may diminish the additional benefits of digital interventions. These findings underscore the potential value of digital health care for individuals with prepregnancy overweight or obesity, as effective weight control in these populations is crucial. Excess weight not only increases the risk of pregnancy complications but also poses significant long-term health risks for their children [66].

The findings did not align with our hypothesis that digital health interventions would benefit participants experiencing insufficient GWG. This discrepancy may stem from the complex causes of inadequate weight gain during pregnancy, which include factors such as prepregnancy anemia, gestational diabetes, unhealthy lifestyle behaviors (eg, substance abuse), parity, and crowded living conditions [67]. Addressing these issues requires more than education on healthy diets, self-monitoring, or food supplementation alone. Additionally, 4 of the 9 studies on insufficient GWG focused on women with overweight and obese, with the digital health interventions targeting weight loss rather than promoting weight gain. This lack of individualized weight management strategies may have limited the effectiveness of the interventions in supporting appropriate weight gain for women below the recommended range.

In our study, while digital health interventions did not address insufficient GWG, they were associated with a reduction in miscarriage and preterm birth rates, as well as an increase in the prevalence of infants who are SGA. The reduction in miscarriage aligns with findings from Victa et al [68], which demonstrated that telemedicine monitoring during pregnancy significantly lowers miscarriage risk by enabling early detection of potential complications and timely medical intervention. In addition, the decreased risk of preterm birth and neonatal composite outcomes (P<.05) is consistent with the study by Guo et al [15], which highlighted the positive impact of digital health interventions on neonatal health.

Although insufficient GWG is known to increase the risk of miscarriage and preterm birth, our findings suggest that digital interventions may mitigate these risks through mechanisms beyond weight control. These interventions likely promote comprehensive health management by emphasizing nutritional intake, lifestyle improvements, and self-management efficacy, thereby improving maternal health and reducing risks associated with miscarriage and preterm labor, even without significantly addressing inadequate weight gain [69].

Further supporting this perspective, our analysis revealed a significant reduction in composite neonatal complications (P<.05), including conditions such as hypoglycemia, jaundice, and acute respiratory distress syndrome. By focusing on combined outcomes, the analysis provides a holistic view of the intervention’s benefits, which aligns more closely with real-world clinical scenarios and underscores the broader potential of digital health interventions to improve maternal and neonatal health.

Pooling data revealed an unexpected correlation between digital health interventions and a higher incidence of infants who are SGA. This outcome may be attributed to the characteristics of the study population, as 9 out of the 11 studies on SGA involved participants with high-risk factors—5 focused on overweight and obesity, 2 on GDM, and 2 on other conditions. Interventions targeting weight control likely resulted in overly strict weight gain or dietary restrictions, increasing the energy gap [70] and potentially impairing placental development [71]. In addition, including participants already at substantial risk for SGA-related conditions may have amplified the negative effects of these interventions. Moreover, the inclusion of participants already at substantial risk for SGA-related conditions may have amplified the observed effects, complicating the interpretation of the actual impact of digital health interventions on SGA. While subgroup analyses could provide more nuanced insights, the variability and limited availability of stratified BMI data across studies hindered such analyses. Future studies should prioritize detailed BMI stratification and reporting to better elucidate the differential impacts of digital health interventions on SGA incidence.

Despite this, digital health interventions offer significant benefits for women and newborns through multiple mechanisms. They provide evidence-based, tailored educational resources that address women’s specific needs [72,73] while encouraging active involvement in health monitoring. These tools enable women to track food intake, physical activity, and physical parameters such as weight changes and blood pressure [74], with positive feedback reinforcing their ability to manage their health effectively. Moreover, some intervention platforms facilitate online communication with skilled clinical and nursing personnel, reducing pregnancy-related stress and enhancing health literacy [75]. These multifaceted approaches contribute to improved maternal and neonatal outcomes, even when specific challenges persist.

The strengths of this study encompass its inclusive population, as it considered all pregnant women rather than concentrating exclusively on specific subgroups such as those with diabetes or obesity, thereby enhancing the generalizability of the findings. In addition, we used a wide range of outcome measures, providing a comprehensive understanding of the impact of digital health interventions on maternal and neonatal health. Furthermore, our study incorporated the most recent literature, ensuring relevance to current clinical practice.

However, the study also has limitations. First, we expanded the search terms to include a broader range of eligible studies, thus enabling a more comprehensive investigation. Nevertheless, the quantity of published RCTs was restricted, indicating that the findings of our study necessitate careful interpretation and require further validation through supplementary research. Second, the follow-up period for the majority of the included studies concluded at birth, indicating a need for further investigation into the long-term effects on women and children to understand the implications of our findings fully. Third, the large sample size in the study by Skalecki et al [60], which focused on telemonitoring fetal movement, may influence the overall results. While a larger sample size is often beneficial for increasing the statistical power of a study, it may also introduce a bias toward the generalizability of the findings, especially in the context of the intervention types and outcome measures. A more specific analysis of how large sample sizes in studies with different intervention types could affect the overall results and the generalizability of the findings would provide a more comprehensive perspective. Fourth, the included studies encompassed a heterogeneous population of pregnant women with varying BMI categories (eg, underweight, normal weight, overweight, and obesity). However, some studies did not stratify participants by prepregnancy BMI or provide detailed subgroup analyses. This lack of consistent BMI reporting limits our ability to interpret the differential effects of digital health interventions across BMI subgroups. Future studies with standardized BMI stratification and detailed subgroup reporting are essential to delineate these effects better. Fifth, while the reduction in GWG observed in our study has general clinical significance due to the known risks of excessive weight gain, the lack of specific data on other contextual factors, such as dietary habits, physical activity levels, and socioeconomic status, may affect the interpretation of the results. These factors could mediate the relationship between digital health interventions and GWG, highlighting the need for more comprehensive datasets in future research.

Further studies should investigate the potential differential impacts of these interventions across various populations, considering factors such as age, ethnicity, and health conditions, and determine the most effective digital intervention tailored to participants’ backgrounds.

Digital health interventions facilitate maternal weight gain, which later influences neonatal health outcomes by decreasing complications such as miscarriages, preterm birth, and combined neonatal complications.